PMML Model Appender

This node combines two PMML document fragments, where one contains a model and the other transformations. In the workflow the process of documenting the transformations and creating the model are separated, so this node can combine them. References to transformed fields in the model are changed to point to the actual derived fields, which are denoted by stars (*) as postfix of the field name. When adding the transformations it is ensured that only fields from the data dictionary are present in the mining schema of the model. When a model is learned on transformed data, it may have transformed fields in the data dictionary and therefore also in the mining schema. Those fields can be safely removed because they are also present in the transformation dictionary of the second input. If a model is learned with a transformed field being the target, the field is removed from the transformation dictionary and a warning is given.

Views

This node has no views

Workflows

• 01_Example_for_Using_PMML_for_Transformation_and_PredictionKNIME Hub
• 01_Model_Selection_SampledKNIME Hub
• 3. My First ANNKNIME Hub
• 3. My First ANNKNIME Hub
• AutoML - Preview (PMML)KNIME Hub
• Show all 7 workflows

Links

• No links available

Developers